fuel cells systems analysis - energy
TRANSCRIPT
Fuel Cells Systems Analysis
R. K. Ahluwalia, X. Wang, K. Tajiri and R. Kumar
2010 DOE Hydrogen Program Review
Washington, DC
June 8-11, 2010
This presentation does not contain any proprietary, confidential, or otherwise restricted information.
Project ID: FC017
2
Overview
Timeline Start date: Oct 2003 End date: Open Percent complete: NA
BarriersB. CostC. PerformanceE. System Thermal and Water
ManagementF. Air ManagementJ. Startup and Shut-down Time,
Energy/Transient Operation
Budget FY10 funding: $650K (+$300K)
DOE share: 100% FY09 funding: $600K (+$300K)
Partners/Interactions Honeywell CEM+TWM projects DTI, TIAX 3M, Emprise, PermaPure ISO-TC192 WG12, HNEI,
JARI, LANL IEA Annexes 17 and 20 FreedomCAR fuel cell tech team
This project addresses system, stack and air management targets for efficiency, power density, specific power, transient response time, cold start-up time, start up and shut down energy
3
Objectives
Develop a validated system model and use it to assess design-point, part-load and dynamic performance of automotive and stationary fuel cell systems. Support DOE in setting technical targets and directing
component development Establish metrics for gauging progress of R&D projects Provide data and specifications to DOE projects on
high-volume manufacturing cost estimation
4
Approach
Develop, document & make available versatile system design and analysis tools. GCtool: Stand-alone code on PC platform GCtool-PSAT: Drive-cycle analysis of hybrid fuel cell
vehicles
Validate the models against data obtained in laboratory and at Argonne’s Fuel Cell Test Facility. Collaborate with external organizations
Apply models to issues of current interest. Work with FreedomCAR Technical Teams Work with DOE contractors as requested by DOE
5
Collaborations
Air Management Honeywell Turbo TechnologiesStack 3M, NuveraWater Management Honeywell Aerospace, Emprise, PermaPureThermal Management Honeywell Thermal SystemsFuel Economy ANL (PSAT)H2 Impurities HNEI, JARI, LANL, ISO-TC-192 WGSystem Cost DTI, TIAXDissemination IEA Annex 22 and 25
– Argonne develops the fuel cell system configuration, determines performance, identifies and sizes components, and provides this information to TIAX for high-volume manufacturing cost estimation
6
Technical Accomplishments1. System analysis to update the status of technology Stack: Determined the performance of NSTFC stacks with 0.15
mg/cm2 Pt loading and 20-µm membrane Air Management: Evaluated the performance of Honeywell’s
compressor-expander-motor module for 1.5-atm operation Fuel Management: Evaluated the performance of parallel ejector-
pump hybrids Water Management: Constructed performance maps for planar
membrane humidifiers Thermal Management: Collaborated with Honeywell to evaluate
performance of microchannel automotive radiators Drive Cycle Simulations: GCtool-PSAT simulations for fuel economy
of hybrid FCEVs Cost: Assisting TIAX in projecting cost of Argonne FCS-2010 at high
volume manufacturing2. Hydrogen impurity effects (Backup Slides) Conducting dynamic simulations to projected combined effect of H2
impurities at ISO specs Providing modeling support to ISO-TC192 WG-12 efforts
7
Argonne 2010 FCS Configuration
2010 FCSMEA- 3M NSTFC MEA- 20-µm 3M membrane- 0.05(a)/0.1(c) mg/cm2 Pt- Metal bipolar plates
AMS- Honeywell CEMM- Air-cooled motor/AFB
WMS- Cathode MH with precooler
TMS- Advanced 40-fpi
microchannel fins
FMS- Parallel ejector-pump hybrid
S1 – Pressurized FCS, 2.5 atm stack inlet pressure at rated power S2 – Low-pressure FCS, 1.5 atm stack inlet pressure at rated power S3 – S2 without cathode humidifier, HT coolant in pre-cooler
8
Air Management System Determined performance of Honeywell’s compressor-expander-
motor module (CEMM) originally designed for 2.5-atm peak P (S1) Comparable component efficiencies for S1 and S2 which may
improve with redesign for higher rpm
Mixed axial flow compressor
Variable nozzle turbine (VNT)
3-phase brushless DC motor, liquid and air cooled
Motor controller, liquid cooled
Air foil bearing (AFB)
0
20
40
60
80
100
0 20 40 60 80 100Mass Flow Rate (%)
Effic
ienc
y (%
) CompressorMotor
Expander
Motor Controller
30
50
70
90
110
Shaf
t Spe
ed (k
rpm
)
S1
S2
1.0
1.5
2.0
2.5
Dis
char
ge P
(atm
) S1
S2
9
Performance of Integrated CEM Module Maximum turndown may be limited by compressor surge for shaft
speeds less than 45 krpm At rated power, the CEMM consumes ~10 kWe in S1, <6 kWe in S2 CEMM min. power between 270-400 We, determined by AFB durability
1
10
100
30 40 50 60 70 80 90 100 110Shaft Speed (krpm)
Turn
dow
n S1
S2
Surge Limit
0
2
4
6
8
10
0 10 20 30 40 50 60 70 80FCS Net Power (kWe)
CEM
Pow
er (k
We)
S2
S1
10
Reference Stack with 3M’s NSTF Catalysts 3M’s single cell data with 0.1(a)/0.15(c) mg/cm2 Pt, ternary PtCoMn
catalyst, and 35-µm 850 EW membrane 3M’s single cell data with 0.05(a)/0.1(c) mg/cm2 Pt, ternary PtCoMn
catalyst, and 20-µm 850 EW membrane ECSA, specific activity, short and crossover currents and HFR data
from CV, EIS and H2/air cells at 0.9 V, 70-120oC, 20-100% RH Determined optimum combination of stack temperature and inlet RH
as a function of pressure, stoichiometry and MEA parameters
FCS S1 S2 S3
Stack P, atm 2.5 1.5 1.5
Stack T, °C 85 75 65
Dew Point T (c), °C 64 61 22
Dew Point T (a), °C 59 53 22
0.5
0.6
0.7
0.8
0.9
0.0 0.3 0.6 0.9 1.2 1.5 1.8Current Density (A/cm2)
Cel
l Vol
tage
(V)
0.0
0.3
0.6
0.9
1.2
Pow
er D
ensi
ty (W
/cm
2 )S2
S1
S3
S1
S2
S3
Cell Voltage
Power Density
11
Stack Performance – Effect of Pt Loading Reference 2009 S1 system: 0.25 mg-Pt/cm2, 35-µm membrane Reference 2010 S1 system: 0.15 mg-Pt/cm2, 20-µm membrane 30-45% projected reduction in Pt content because of lower Pt loading
and thinner membrane, 0.12-0.30 g-Pt/kWe(net)
500
1000
1500
mW
/cm
2
Power Density at Rated Power
0.15 mg-Pt/cm2
20-μm membrane
0.25 mg-Pt/cm2
35-μm membrane
0.1
0.2
0.3
0.4
40 45 50System Efficiency (%)
g-Pt
/kW
e
Pt Content on System Basis
0.15 mg-Pt/cm2
20-μm membrane
0.25 mg-Pt/cm2
35-μm membrane
12
Stack Performance – Effect of Operating Pressure S1 has higher power density and lower Pt content even though cell V
has to be 25-35 mV higher to compensate for larger parasitic power
550
600
650
700
750
mV
S1
S2
Cell Voltage at Rated Power
500
1000
1500
mW
/cm
2 S1
S2
Power Density at Rated Power
0.1
0.2
0.3
40 45 50System Efficiency (%)
g-Pt
/kW
e
S1
S2
Pt Content on System Basis
13
Fuel Management System Parallel ejector-pump hybrid Ejector performance
– Motive (p) gas: cH2, 15-atm maximum P, MW=2– Suction (s) gas: H2 with water, 1-1.5 atm, 75oC, MW=3-7– Lift pressure: <3 psi; Recycle ratio: ms/mp = 2-5 – Blower flow rate(%): 100-Entrainment(%)
0
20
40
60
80
100
120
0 20 40 60 80 100Power (%)
Entr
ainm
ent (
%)
0
3
6
9
12
15
18
Mot
ive
Gas
Pre
ssur
e (a
tm)
Ejector OnlyBlower OnlyEjector
+Blower
M > 1
M < 1
14
Alternate Arrangements A blower is always needed if the FCS has a turndown >3 with single
ejector, >4 with variable geometry ejector, and > 5 with dual ejectors
Dual Ejectors + Blower Variable Geometry Ejector + Blower
Blower None Single Variable Geometry Dual
Flow Rate L/s 20 5.6 4.5 3.1Pressure Head psi 3 0.8 0.65 0.4Power W 400 35 25 10
Ejector Arrangement
0
20
40
60
80
100
120
0 20 40 60 80 100Power (%)
Entr
ainm
ent (
%)
0
3
6
9
12
15
18
Mot
ive
Gas
Pre
ssur
e (a
tm)
Ejector 1 OnlyEjector 2 Only
Blower Only
E2+ B
M < 1
M > 10
20
40
60
80
100
120
0 20 40 60 80 100Power (%)
Entr
ainm
ent,
Thro
at A
rea
(%)
Ejector OnlyBlower Only Ejector+
BlowerEntrainment
Throat Area
15
Water Management System Analyzed Honeywell and PermaPure data for full-scale (FS), half-scale
(HS) and 1/10th sub-scale (SS) membrane humidifiers Permeance (κm) for the units (not local values) can be represented in
terms of the dry-air outlet RH. – κm defined similarly as LMTD for heat transfer– κm also depends on temperatures
Effectiveness (εm) can be represented in terms of NMTU (number of mass transfer units)
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0 2 4 6 8 10 12 14 16NMTU
Effe
ctiv
enes
s
80C FS 90C FS80C HS 90C HS100% Flow SS 75% Flow SS25% Flow SS
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0 20 40 60 80 100
RH of Humidified Air (%)
Perm
eanc
e (m
ol/m
2 .s.a
tm)
80C FS 90C FS
80C HS 90C HS
100% Flow SS 75% Flow SS
25% Flow SS
16
Water Mass Transfer Flux Model results for mass transfer from saturated wet stream Optimum dry-air inlet temperature (Tm) for maximum flux
– Flux decreases for Td > Tm because of low uptake (too dry)– Flux decreases for Td < Tm as humidified air approaches saturation
Flux is also a strong function of the membrane thickness, temperature of the wet air and operating pressures
0
1
2
3
4
5
6
7
8
45 55 65 75 85 95
Dry Air Inlet Temperature (oC)
Wat
er F
lux
(g/m
2 .s)
Membrane Thickness = 20 µmP = 2.5 atmWet Air T = 80oCWet Air RH = 100%
Approach TDP
5oC
10oC
15oC
35oC
17
Thermal Management System Heat rejection more difficult in FCVs than ICEVs
– Higher efficiency, but similar or higher heat loads but smaller ∆T– Additional FCS and electric drivetrain components requiring cooling
to lower temperatures: stacked A/C condenser, LTR and HTR– Heat rejection in FC powertrains is most challenging while driving at
55 mph on 6.5% grade with 600 kg payload– Allow stack temperature to rise when heat rejection is difficult– Expander needed even in S2 and S3
Rated Power kWe 80 80 80 80 80 80 80 80 80System Efficiency % 50 45 40 50 45 40 50 45 40Stack Pressure atm 2.5 2.5 2.5 1.5 1.5 1.5 1.5 1.5 1.5Stack Temperature oC 85 85 85 75 75 75 65 65 65Heat RejectionDemand Power kWe 61.5 61.5 61.5 61.5 61.5 61.5 61.5 61.5 61.5Cell Voltage mV 741 681 634 732 682 649 734 688 657Stack Pressure atm 2.4 2.4 2.3 2.1 2.1 2 2.1 2.1 2Stack Temperature oC 95 95 92 95 95 92 80 80 75HT Radiator Heat Load kW 49.5 59.5 68.3 49.3 57.6 63.5 49.4 56.7 62.2LT Radiator Heat Load kW 16.2 16.5 16.2 13.6 13.8 13.5 13.3 13.5 13.3
S1 S2 S3
18
Microchannel Radiator Derived f and j factors from Honeywell data for 9”x9”x1.3” subscale
and 27.6”x17.7”x1.3” full-scale radiators – 18 and 24 fpi louver and 40 and 50 fpi microchannel fins
FCS radiator can be more compact with 40-fpi microchannel than with standard automotive18-fpi louver fins
FC powertrains will likely be derated at Ta > 40oC since the fan power doubles for every 5oC increase in ambient temperature (Ta)
0
200
400
600
800
10 20 30 40 50 60 70 80 90 100HT Radiator Depth (mm)
Fan
Pum
ping
Pow
er (W
)
45°C (1.25)
MC-40Stack T: 95oCHeat Load: 56 kWVehicle Speed: 55 mph
Ambient T (A/Aref)
45°C (1.5)
40°C (1.25)
40°C (1.5)
19
Heat Rejection vs. Stack Efficiency at Rated Power
0
200
400
600
800
1000
60 70 80 90 100
Stack Temperature (oC)
Min
imum
Fan
Pum
ping
Pow
er (W
)
40%
45%50%
FCS Efficiency
MC-40A/Aref = 1.25Ambient T = 40oC
25
50
75
100
60 70 80 90 100
Stack Temperature (oC)
HT
Rad
iato
r Dep
th (m
m)
40%45%50%
FCS Efficiency
MC-40A/Aref = 1.25Ambient T = 40oC
At low stack temperatures, the minimum fan power increases nonlinearly with heat load (i.e., FCS efficiency at rated power)
For given fan power, 40% FCS (S2) must be capable of operating at 5-10oC higher stack temperature than 50% FCS
From the standpoint of heat rejection, 40% FCS may be acceptable if it can be operated at 95oC (S3 not OK)
20
System Performance - SummaryLowering rated-power efficiencyfrom 50% to 40% decreases peak efficiency by <1% fuel economy by 4% in battery-
charging mode (BCM) & 7% in load-following mode (LFM)
and system Pt content by 50%No. of start-stops on UDDS: 58 inBCM, 4 in LFM2
PSAT drive-cycle simulation results for a mid-size hybrid vehicle
48
49
50
51
52
53
54
55
56
BCM LFMUDDS
FCS
Effic
ienc
y (%
)
S2-3 (40%) S2-2 (45%) S2-1 (50%)
BCM LFMHWFET
BCM LFMLA92
70
72
74
76
78
80
82
84
86
88
BCM LFMUDDS
Fuel
Eco
nom
y (m
pgge
)
S2-3 (40%) S2-2 (45%) S2-1 (50%)
BCM LFMHWFET
50
52
54
56
58
60
62
64
66
68
BCM LFMLA92
BCM LFMCombined
20
25
30
35
40
45
50
55
60
0 10 20 30 40 50 60 70 80Power (kWe)
Effic
ienc
y (%
)
50%
45%
40%
21
Future Work1. Systems Analysis Support DOE/FreedomCAR development effort at system,
component, and phenomenological levels Collaborate with 3M to develop durability models for NSTFC
electrode structures Continue cooperation with Honeywell and others to validate air, fuel,
thermal, and water management models System optimization for cost, performance, and durability Drive cycle simulations for durability enhancement Alternate membrane, catalyst structures, and system configurations Support DTI and TIAX in high-volume manufacturing cost projections 2. Hydrogen Quality Validate impurity models against U.S. and JARI data Project effects of proposed standards on stack performance Provide modeling support to ISO-TC192 WG-12 and the Codes and
Standards Technical Team
28
1. SMR-PSA may be the pathway for producing H2 in the early stages of hydrogen economy
CO and N2 levels determine the H2 recovery from PSA and influence the cost of producing H2
Of the two impurities, CO degrades the cell performance more but accumulation of N2 accumulation can significantly dilute the H2 concentration
Our simulations show that N2 crossover from air is the main source of N2 buildup in the anode recycle stream and determines the purge schedule
2. H2S and NH3 are extremely harmful to PEFC efficiency but are easily removed in PSA beds
Dynamic simulations show that NH3 cannot accumulate in the anode recycle stream
Dynamic Behavior of Fuel Hydrogen Impurities in Polymer Electrolyte Fuel Cells
29
Hydrogen Purification Drivers (PSA)
Relatively easier to remove
0.05 ppmStrongHalogenates
Relatively easier to remove
Dew PointDew Point5 ppmStrongWater (H2O)
Relatively easier to remove
0.004 ppmStrongTotal Sulfur
Relatively easier to remove
Low ppmLow ppm0.1 ppmStrongAmmonia
Relatively easier to remove
25000.5%5000.1 %2 ppm(incl CH4)
|Total HC’s
Relatively easier to remove
9000015 -18%8500015-17%2 ppm|Carbon Dioxide (CO2)
Impacts PSA recovery & Capital Cost
150000.5 – 3%100000.5 – 2%2 ppm(incl THC)
Methane (CH4)
Impacts PSA recovery & Capital Cost
2000000.1-4%500000.1 -1 %0.2 ppm|Carbon Monoxide (CO)
Impacts PSA recovery & Capital Cost
101000 ppm380034-38%100 ppm(total inert)
|Nitrogen (N2)
Impacts PSA recovery & Capital Cost
5500 ppm5500 ppm100 ppm(total inert)
|Argon (Ar)
Impacts PSA recovery & Capital Cost
--1050 ppm5 ppm|Oxygen (O2)
Impacts PSA recovery & Capital Cost
75-80%40-45%99.99%WeakHydrogen (H2)
NOT POSSIBLE5500 ppm5500 ppm100 ppm(total inert)
ZeroHelium (He)
OVERALLEFFECT
Purification Ratio for SMR
SMRMol %
Purification Ratio for ATR
ATRMol %
ISO TC 197 WG 12 (14687)
Draft Spec
Adsorption Force
Species
Relatively easier to remove
0.05 ppmStrongHalogenates
Relatively easier to remove
Dew PointDew Point5 ppmStrongWater (H2O)
Relatively easier to remove
0.004 ppmStrongTotal Sulfur
Relatively easier to remove
Low ppmLow ppm0.1 ppmStrongAmmonia
Relatively easier to remove
25000.5%5000.1 %2 ppm(incl CH4)
|Total HC’s
Relatively easier to remove
9000015 -18%8500015-17%2 ppm|Carbon Dioxide (CO2)
Impacts PSA recovery & Capital Cost
150000.5 – 3%100000.5 – 2%2 ppm(incl THC)
Methane (CH4)
Impacts PSA recovery & Capital Cost
2000000.1-4%500000.1 -1 %0.2 ppm|Carbon Monoxide (CO)
Impacts PSA recovery & Capital Cost
101000 ppm380034-38%100 ppm(total inert)
|Nitrogen (N2)
Impacts PSA recovery & Capital Cost
5500 ppm5500 ppm100 ppm(total inert)
|Argon (Ar)
Impacts PSA recovery & Capital Cost
--1050 ppm5 ppm|Oxygen (O2)
Impacts PSA recovery & Capital Cost
75-80%40-45%99.99%WeakHydrogen (H2)
NOT POSSIBLE5500 ppm5500 ppm100 ppm(total inert)
ZeroHelium (He)
OVERALLEFFECT
Purification Ratio for SMR
SMRMol %
Purification Ratio for ATR
ATRMol %
ISO TC 197 WG 12 (14687)
Draft Spec
Adsorption Force
Species
DOE H2QWG Draft Roadmap. Courtesy, Bhaskar Balasubramanian (Chevron)
30
CO/CO2 Poisoning Model
Hydrogen Oxidation Reaction
– H2 + 2M ↔ 2M-H (Dissociative Adsorption)
– M-H → M + H+ + e- (Electrochemical Oxidation)
CO Poisoning of Pt
– CO + M ↔ M-CO (Associative Adsorption on Linear Sites)
– CO2 + 2M-H → M + M-CO + H2O (Reverse Water-Gas Shift)
– M-CO + H2O → M + CO2 + 2H+ + 2e- (Electrochemical Oxidation)
Reactions with Oxygen
– M-CO + ½ O2 → M + CO2 (CO Oxidation)
– 2M-H + ½ O2 → 2M + H2O (H2 Oxidation)
31
H2S Poisoning Model
Sequential Sorption of H2S on Pt
1. nM + H2S ↔ Mn-H2S (Associative chemisorption)
2. Mn-H2S ↔ MnS + 2H+ + 2e- (Electrochemical reaction)
3. MnS + 3H2O → nM + SO3 + 6H+ + 6e- (Electrochemical oxidation)
Multi-site sorption of H2S, n is a function of total sulfur coverage (θS)
– n →1 as θS → 1, n → N as θS → 0
Near OC, MnS can re-convert to Mn-H2S (E2 = 0.14 V), and H2S can desorb for partial recovery
At a high anode overpotential (E3 = 0.89 V), MnS can oxidize to SO3, SO3 assumed completely soluble in water and removed from the system
32
Experimental Observations Reversible at low concentrations but may be irreversible at high
concentrations (Uribe 2002) CV traces suggest that NH3 is not significantly adsorbed on the
electrocatalyst (Uribe 2002 & Soto 2003). AC impedance spectroscopy data indicate that increased
membrane resistance cannot account for the observed overpotentials due to NH3 impurity (Soto 2003).
Model Transient stack model with steady-state option NH3 uptake in ionomer modeled as a reversible absorption-
desorption reaction Same approach for NH3 uptake in membrane except it is exposed
to both anode and cathode streams Effect of NH4
+ on conductivity from experimental data Postulated effect of NH4
+ on ORR kinetics
NH3 Poisoning Model
33
HCHO/HCOOH/C7H8 Poisoning Model
Hydrogen Oxidation Reaction
– H2 + 2M ↔ 2M-H (Dissociative Adsorption)
– M-H → M + H+ + e- (Electrochemical Oxidation)
Poisoning of Pt
– HCHO + M → M-CO + 2H++ 2e- (Dissociative Adsorption)
– HCOOH + M → M-CO + H2O (Dissociative Adsorption)
– C7H8 + M ↔ M-C7H8 (Associative Adsorption)
Reaction with Hydrogen
– C7H8 + 6M-H → 6M + C7H14 (Hydrogenation to MCH)
34
NH3 Impurity Model Development (JARI Data)
0
30
60
90
120
150
0.25 0.30 0.35 0.40 0.45 0.50
Anode/Cathode Pt Loading (mg/cm2)
DV (m
V)
T, Tdp = 80/77(a)/50(c)oCNH3 Conc. = 10 ppmUtiliization = 17(a)/40(c)%J = 1 A/cm2
Modeled NH3 crossover is a function of uptake and current density
0
10
20
30
40
0.0 0.2 0.4 0.6 0.8 1.0NH3 Concentration (ppm)
DV (m
V)
T, Tdp = 80/77(a)/50(c)oCPt Loading = 0.3 mg/cm2
Utiliization = 17(a)/40(c)%J = 1 A/cm2
Exposure Time = 10 h
35
HCHO/HCOOH Impurity Model Development:JARI Data
0
5
10
15
20
25
0 4 8 12 16 20HCHO Concentration (ppm)
DV (m
V)
T, Tdp = 80/77(a)/50(c)oCPt Loading = 0.3 mg/cm2
Utiliization = 17(a)/40(c)%J = 1 A/cm2
0
5
10
15
20
0 100 200 300 400 500HCOOH Concentration (ppm)
DV (m
V)
T, Tdp = 80/77(a)/50(c)oCPt Loading = 0.3 mg/cm2
Utiliization = 17(a)/40(c)%J = 1 A/cm2
36
Toluene alone: Small ∆V, significant conversion to methyl-cyclohexane Toluene + CO: Larger ∆V, insignificant hydrogenation
Analysis of HNEI Data: Toluene
0.0
0.2
0.4
0.6
0.8
1.0
1.2
0 20 40 60 80 100Time, h
Flow
Rat
e, n
mol
/s
CO In
A/C: 7/7psig, 2/2 Stoich 100/50% RH, H2/Air60oC2 ppm CO1 A/cm2
CO Out
CO2 Out
CO2 In
0
2
4
6
8
10
12
0 20 40 60 80 100Time, h
Flow
Rat
e, n
mol
/s
C7H8 In
A/C: 7/7psig, 2/2 Stoich 100/50% RH, H2/Air60oC20 ppm C7H8
1 A/cm2
C7H8 Out
C7H14 Out
0
100
200
300
400
0 20 40 60 80 100Time, h
DV, m
V
A/C: 7/7psig, 2/2 Stoich 100/50% RH, H2/Air60oC1 A/cm2
20 ppm C7H8
2 ppm CO
2 ppm CO + 20 ppm C7H8
0
2
4
6
8
10
12
0 20 40 60 80 100Time, h
Flow
Rat
e, n
mol
/s
C7H8 In
A/C: 7/7psig, 2/2 Stoich 100/50% RH, H2/Air60oC20 ppm C7H8+2 ppm CO1 A/cm2
C7H8 Out
C7H14 OutCO In CO2 Out
2-ppm CO
20-ppm C7H8
∆V
2-ppm CO +20-ppm C7H8
37
Effect of Inerts (N2, He, CH4) Giner data for N2 permeance as a function of T and membrane
water content (fv)– N2 crossover from cathode air may potentially exceed N2 in H2
He permeance from JARI data: 2.4x10-13 mol/cm.s.kPa at 80oC, 77oC dew point, 1.6 higher than H2 permeance
CH4 and CO2 are relatively impermeable (HNEI, JARI)
0.0E+00
2.0E-14
4.0E-14
6.0E-14
8.0E-14
1.0E-13
1.2E-13
1.4E-13
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35fv
Perm
eanc
e (m
ole/
cm.s
.kPa
)
95oC
80oC
60oC
30oC
38
1. Pressurized stack 87 kWe at 1 A/cm2 with pure H2
50% O2 & 70% per-pass H2 utilization N111 membrane, 0.2(c)/0.1(a) mg/cm2 Pt loading Stack pressure: 2.5/1.9/1.6 atm at 100/25/5% power Inlet RH: 60/80/90% at 100/25/5% power2. Purge criteria Single purge (14 L) equivalent to 2 volumes of anode subsystem
restores gas in anode channels to the H2 fuel specs 3. Definitions Concentrations generally reported as max. values after 30-h
exposure Stack efficiency defined as ratio of DC power produced to the LHV
of H2 consumed (chemically and electrochemically) and purged
Reference Conditions, Assumptions, Definitions
39
Dynamic simulations of Impurity Effects Purge occurs whenever impurity concentration reaches 15%
779
780
781
782
Cel
l Vol
tage
(V) 0.25 A/cm2
645
650
655
660
0 5 10 15 20 25 30Time (h)
Cel
l Vol
tage
(mV) 1 A/cm2
40
Purge Loss and Schedule
At low power, purge interval determined primarily by N2 crossover At high power, buildup of fuel impurities also affects purge interval
0
2
4
6
8
10
12
0 25 50 75 100Stack Power (%)
Purg
e Lo
ss (%
)
0
100
200
300
400
500
600
Purg
e In
terv
al, s
Purge Loss
Purge Interval
Pure H2
ISO H2
Pure H2
ISO H2
CO: 0.2 ppmCO2: 2 ppmN2: 100 ppmHe: 200 ppmNH3: 0.1 ppmHCHO: 10 ppbHCOOH:0.2 ppmC7H8: 2 ppmH2S: 4 ppb
41
Dynamic Pt Coverage
Steady-state coverage of CO & C7H8 functions of P & overpotentials No SS coverage for H2S which competes with CO & C7H8 for vacant sites
0
5
10
15
20
Cov
erag
e (%
)
CO
H2S
100% Power
C7H8
0
2
4
6
0 5 10 15 20 25 30Time (h)
Cov
erag
e (%
) CO
H2S
25% Power
C7H8
42
Buildup of Impurities with Dynamic Purge: N2, He, CO, CO2
0
5
10
15
N2 (
%)
Stack Inlet
Stack Outlet
0
1
2
3
4
He
(%)
Stack Inlet
Stack Outlet
0
150
300
450
600
0 20 40 60 80 100Stack Power (%)
CO
2 (pp
m)
Stack Inlet
Stack Outlet
N2 buildup almost entirely due to N2 crossover from cathode No buildup of CO as it is quantitatively converted to CO2
43
Dynamic Buildup of NH3, H2S and HCHO No buildup of NH3, NH3 crossover increases with current density H2S breakthrough at more than 20% power, >30 h exposure
0.00
0.05
0.10
NH
3 (pp
m) Stack Inlet
Stack Outlet
0
5
10
15
H2S
(ppb
)
Stack Inlet
Stack Outlet
0
5
10
15
20
0 20 40 60 80 100Stack Power (%)
HC
HO
(ppb
)
Stack Inlet
Stack Outlet
Enrichment
)1()1(
2H
ii
ini
outi
i
E
CCE
Φ−Φ−
=
=
44
Dynamic Buildup of HCOOH and C7H8
Slower dissociative adsorption of HCOOH than HCHO Conversion of toluene to methyl-cyclohexane slows at higher power
0
5
10
15
HC
OO
H (p
pm)
Stack Inlet
Stack Outlet
0.0
0.5
1.0
1.5
2.0
C7H
8 (pp
m) Stack Inlet
Stack Outlet
0
100
200
300
400
0 20 40 60 80 100Stack Power (%)
C7H
14 (p
pm)
Stack Inlet
Stack Outlet
45
Simultaneous Effect of Impurities
Losses defined with respect to cell voltage and stack efficiency for pure fuel hydrogen accounting for N2 crossover from cathode
At <30% power, ∆V approaches zero but 0.35 percentage point reduction in stack efficiency due to buildup of impurities
Slow increase in ∆V expected at longer exposure because of H2S uptake
0
5
10
15
0 20 40 60 80 100Stack Power (%)
DV (m
V)
0.00
0.25
0.50
0.75
1.00
Dh (%
)
DV
DhPt Loading: 0.1(a)/0.2(c) mg/cm2
Exposure Time: 30 hTemperature: 80oC
46
Effect of Anode Pt Loading
Substantially higher ∆V and ∆η with 0.05 mg/cm2 Pt (a) for >50% stack power
0
10
20
30
DV (m
V)
0.1 mg/cm2 Pt(a)
0.05 mg/cm2 Pt(a)Exposure Time: 30 hTemperature: 80oC
0.0
0.5
1.0
1.5
2.0
0 20 40 60 80 100Stack Power (%)
Dh (%
)
0.1 mg/cm2 Pt(a)
0.05 mg/cm2 Pt(a)
47
Effect of Stack Temperature
70oC stack temperature, 1.5 bar inlet pressure at rated power
0
10
20
30
DV (m
V)
70oCPt Loading: 0.1(a)/0.2(c) mg/cm2
Exposure Time: 30 h
80oC
0.0
0.5
1.0
1.5
2.0
0 20 40 60 80 100Stack Power (%)
Dh (%
)
70oC
80oC
48
Sensitivity Analysis Results for 100% stack power, ISO max impurity specs with one
species at 2X and 0.5 X concentration
0
5
10
15
20
25
CO CO2 N2 He NH3 HCHO HCOOH C7H8 H2S
DV (m
V)
Pt Loading: 0.1(a)/0.2(c) mg/cm2
Exposure Time: 30 hTemperature: 80oC
0.0
0.5
1.0
1.5
2.0
CO CO2 N2 He NH3 HCHO HCOOH C7H8 H2S
Dh (%
)
49
Drive Cycle Simulations Multiple repeats of warm FUDS (1372 s) and FHDS (740 s) for 30 h Insignificant differences in VI curve for the last FUDS-FHDS cycle
0
1
2
3
4
5
Cov
erag
e (%
) CO
H2SC7H8
50
51
52
53
54
55
0 5 10 15 20 25 30Time (h)
Effic
ienc
y (%
)
Pure H2
ISO H2
Pt Loading: 0.1(a)/0.2(c) mg/cm2
Exposure Time: 30 hTemperature: 80oC
50
Summary and Conclusions N2 crossover from air is the main source of N2 buildup in the anode
recycle stream and determines the purge schedule Dynamic simulations show that NH3 cannot accumulate in the anode
recycle stream– Significant crossover of NH3 to cathode air– Cumulative degradation in performance due to H2S
Cyclic buildup of formaldehyde, formic acid and methyl-cyclohexane Critical data needs
– CO/CO2: H2 pump data for independent verification of chemical vs. electrochemical CO oxidation
– NH3 crossover as function of current density, NH3 isotherms as functions of RH and T
– H2S: Isolation of anode vs. cathode overpotentials– Effect of T on behavior of HCHO, HCOOH and C7H8